WebSVN – Kolibri OS – Diff – /programs/demos/view3ds/3dmath.inc


x3d equ 0
y3d equ 2
z3d equ 4
vec_x equ 0
vec_y equ 4
vec_z equ 8
 
if 0  ; Ext >= SSE3
calc_bounding_box:
; in:
;      xmm0  -  normal vector of ray
;      xmm1  -  light origin
; out:
;      eax - axis aligned bounding boxes bit mask
 
       .rmx     equ [ebp-36]
       .nray    equ [ebp-64]
       .origin  equ [ebp-80]
       .dirfrac equ [ebp-96]
       .nrayr   equ [ebp-112]
       .originr equ [ebp-128]
       .tmin    equ [ebp-132]
       .tmax    equ [ebp-136]
 
 
       push    ebp
       mov     ebp,esp
       and     ebp,-16
       sub     esp,160
 
       movss     xmm5,[rsscale]
       shufps    xmm5,xmm1,0
       movd      xmm2,[vect_x]
       punpcklwd xmm2,[the_zero]
       cvtdq2ps  xmm2,xmm2
       subps     xmm1,xmm2
       movaps    .origin,xmm1
       mulps     xmm0,xmm5
       movaps    .nray,xmm0
 
       mov     esi,matrix
       lea     edi,.rmx
       call    reverse_mx_3x3
 
;  in:  esi - ptr to points(normals], each point(normal) coeficient as dword
;       edi - ptr to rotated points(normals)
;       ebx - ptr to 3x3 (9 dwords, 36 bytes) rotation matrix
;       ecx - number of points(normals)
 
    ; reverse transform
       lea    esi,.nray
       lea    edi,.nrayr
       lea    ebx,.rmx
       mov    ecx,1
       call   rotary
 
       lea    esi,.origin
       lea    edi,.originr
       lea    ebx,.rmx
       mov    ecx,1
       call   rotary
 
       xor      ecx,ecx
       mov      ebx,aabb1
       xor      eax,eax
       rcpps    xmm7,.nrayr
       movaps   .dirfrac,xmm7
 
    .nx_aabb:
       movaps   xmm5,[ebx]
       movaps   xmm6,[ebx]
       minps    xmm5,[the_zero]
       maxps    xmm6,[the_zero]
 ;  xmm5 - lb corner of AABB with minimal coordinates
 ;  xmm6 - rt cor.   of AABB wit maximum coords
       subps    xmm5,.originr
       subps    xmm6,.originr
       mulps    xmm5,.dirfrac   ; xmm5 - tx1, ty1
       mulps    xmm6,.dirfrac   ; xmm6 - tx2, ty2
       movaps   xmm1,xmm6
       movaps   xmm2,xmm6
 
 
       minps    xmm1,xmm5
       maxps    xmm2,xmm5
 
       movaps   xmm5,xmm1
       movaps   xmm6,xmm2
       shufps   xmm5,xmm5,11100001b
       shufps   xmm6,xmm6,11100001b
       maxss    xmm1,xmm5  ;t min
       minss    xmm2,xmm6  ;t max
       comiss   xmm2,xmm1
       jb       .no_inter
    .yes:
       bts      eax,ecx
     .no_inter:
       add      ebx,16
       inc      ecx
       cmp      ecx,8
       jne      .nx_aabb
 
; out: eax - bit mask
       add      esp,160
       pop      ebp
ret
end if
 
reverse_mx_3x3:
; esi - source matrix
; edi - desired reversed matrix
 
  push  ebp
  mov   ebp,esp
  sub   esp,4
  .det  equ  ebp-4
 
  fninit
  fld  dword[esi]
  fmul dword[esi+16]
  fmul dword[esi+32]
  fld  dword[esi+12]
  fmul dword[esi+28]
  fmul dword[esi+8]
  faddp
  fld  dword[esi+24]
  fmul dword[esi+4]
  fmul dword[esi+20]
  faddp
  fld  dword[esi]
  fmul dword[esi+28]
  fmul dword[esi+20]
  fchs
  faddp
  fld  dword[esi+24]
  fmul dword[esi+16]
  fmul dword[esi+8]
  fchs
  faddp
  fld  dword[esi+12]
  fmul dword[esi+4]
  fmul dword[esi+32]
  fchs
  faddp
  fstp dword[.det]
  cmp  dword[.det],0
  jne  @f
  int3
 @@:
 ; fld1
 ; fdiv dword[.det]
 ; fstp dword[.det]
 
  fld  dword[esi+16]
  fmul dword[esi+32]
  fld  dword[esi+20]
  fmul dword[esi+28]
  fchs
  faddp
  fdiv dword[.det]
  fstp dword[edi]
 
  fld  dword[esi+8]
  fmul dword[esi+28]
  fld  dword[esi+4]
  fmul dword[esi+32]
  fchs
  faddp
  fdiv dword[.det]
  fstp dword[edi+4]
 
  fld  dword[esi+4]
  fmul dword[esi+20]
  fld  dword[esi+8]
  fmul dword[esi+16]
  fchs
  faddp
  fdiv dword[.det]
  fstp dword[edi+8]
 
  fld  dword[esi+20]
  fmul dword[esi+24]
  fld  dword[esi+12]
  fmul dword[esi+32]
  fchs
  faddp
  fdiv dword[.det]
  fstp dword[edi+12]
 
  fld  dword[esi]
  fmul dword[esi+32]
  fld  dword[esi+8]
  fmul dword[esi+24]
  fchs
  faddp
  fdiv dword[.det]
  fstp dword[edi+16]
 
  fld  dword[esi+8]
  fmul dword[esi+12]
  fld  dword[esi]
  fmul dword[esi+20]
  fchs
  faddp
  fdiv dword[.det]
  fstp dword[edi+20]
 
  fld  dword[esi+12]
  fmul dword[esi+28]
  fld  dword[esi+16]
  fmul dword[esi+24]
  fchs
  faddp
  fdiv dword[.det]
  fstp dword[edi+24]
 
  fld  dword[esi+4]
  fmul dword[esi+24]
  fld  dword[esi]
  fmul dword[esi+28]
  fchs
  faddp
  fdiv dword[.det]
  fstp dword[edi+28]
 
  fld  dword[esi]
  fmul dword[esi+16]
  fld  dword[esi+4]
  fmul dword[esi+12]
  fchs
  faddp
  fdiv dword[.det]
  fstp dword[edi+32]
 
 
  mov  esp,ebp
  pop  ebp
ret
; 3d point - triple integer word coordinate
; vector   - triple float dword coordinate
;----------------------in: --------------------------------
;------------------------ esi - pointer to 1st 3d point ---
;------------------------ edi - pointer to 2nd 3d point ---
;------------------------ ebx - pointer to result vector --
;---------------------- out : none ------------------------
 
make_vector_r:
   if Ext < SSE2
        fninit
        fld     dword[edi]                ;edi+x3d
        fsub    dword[esi]                ;esi+x3d
        fstp    dword[ebx+vec_x]
 
        fld     dword[edi+4]
        fsub    dword[esi+4]
        fstp    dword[ebx+vec_y]
 
        fld     dword[edi+8]
        fsub    dword[esi+8]
        fstp    dword[ebx+vec_z]
    else
        movups  xmm0,[esi]
        movups  xmm1,[edi]
        subps   xmm1,xmm0
        movlps  [ebx],xmm1
        movhlps  xmm1,xmm1
        movss   [ebx+8],xmm1
     end if
 
ret
;---------------------- in: -------------------------------
;--------------------------- esi - pointer to 1st vector --
;--------------------------- edi - pointer to 2nd vector --
;--------------------------- ebx - pointer to result vector
;---------------------- out : none
cross_product:
        fninit
        fld     dword [esi+vec_y]
        fmul    dword [edi+vec_z]
        fld     dword [esi+vec_z]
        fmul    dword [edi+vec_y]
        fsubp   ;st1 ,st
        fstp    dword [ebx+vec_x]
 
        fld     dword [esi+vec_z]
        fmul    dword [edi+vec_x]
        fld     dword [esi+vec_x]
        fmul    dword [edi+vec_z]
        fsubp   ;st1 ,st
        fstp    dword [ebx+vec_y]
 
        fld     dword [esi+vec_x]
        fmul    dword [edi+vec_y]
        fld     dword [esi+vec_y]
        fmul    dword [edi+vec_x]
        fsubp   ;st1 ,st
        fstp    dword [ebx+vec_z]
ret
cross_aligned:
      movaps  xmm0,[esi]
      movaps  xmm1,[esi]
      movaps  xmm2,[edi]
      movaps  xmm3,[edi]
      shufps  xmm0,xmm0,00001001b
      shufps  xmm1,xmm1,00010010b
      shufps  xmm2,xmm2,00010010b
      shufps  xmm3,xmm3,00001001b
      mulps   xmm0,xmm2
      mulps   xmm1,xmm3
      subps   xmm0,xmm1
      movaps  [ebx],xmm0
ret
;----------------------- in: ------------------------------
;---------------------------- edi - pointer to vector -----
;----------------------- out : none
normalize_vector:
if Ext >= SSE2
        movups  xmm0,[edi]
        andps   xmm0,[zero_hgst_dd]
        movups  xmm1,xmm0
        mulps   xmm0,xmm0
        movhlps xmm2,xmm0
        addps   xmm0,xmm2
        movaps  xmm2,xmm0
        shufps  xmm2,xmm2,11100101b
        addps   xmm0,xmm2
        shufps  xmm0,xmm0,0
;        haddps  xmm0,xmm0
;        haddps  xmm0,xmm0
        rsqrtps xmm0,xmm0
        mulps   xmm0,xmm1
        movlps  [edi],xmm0
        movhlps xmm0,xmm0
        movss   [edi+8],xmm0
else
 
        fninit
        fld     dword [edi+vec_x]
        fmul    st, st
        fld     dword [edi+vec_y]
        fmul    st, st
        fld     dword [edi+vec_z]
        fmul    st, st
        faddp   st1, st
        faddp   st1, st
        fsqrt
 
        ftst
        fstsw ax
        sahf
        jnz     @f
 
        fst     dword [edi+vec_x]
        fst     dword [edi+vec_y]
        fstp    dword [edi+vec_z]
        ret
      @@:
        fld st
        fld st
        fdivr dword [edi+vec_x]
        fstp  dword [edi+vec_x]
        fdivr dword [edi+vec_y]
        fstp  dword [edi+vec_y]
        fdivr dword [edi+vec_z]
        fstp  dword [edi+vec_z]
end if
ret
;------------------in: -------------------------
;------------------ esi - pointer to 1st vector
;------------------ edi - pointer to 2nd vector
;------------------out: ------------------------
;------------------ st0 - dot-product
dot_product:
        fninit
;if Ext >=SSE3
;        movups  xmm0,[esi]
;        movups  xmm1,[edi]
;        andps   xmm0,[zero_hgst_dd]
;        mulps   xmm0,xmm1
;        haddps  xmm0,xmm0
;        haddps  xmm0,xmm0
;        movss   [esp-4],xmm0
;        fld     dword[esp-4]
;else
        fld     dword [esi+vec_x]
        fmul    dword [edi+vec_x]
        fld     dword [esi+vec_y]
        fmul    dword [edi+vec_y]
        fld     dword [esi+vec_z]
        fmul    dword [edi+vec_z]
        faddp
        faddp
;end if
ret
 
; DOS version Coded by Mikolaj Felix aka Majuma
; mfelix@polbox.com
; www.majuma.xt.pl
; into FASM translation by Macgub
init_sincos_tab:
.counter   equ  dword [ebp-4]  ; cur angle
 
     push       ebp
     mov        ebp,esp
 
     xor        eax,eax
     push       eax            ; init .counter
     mov        edi,cos_tab
     mov        esi,sin_tab
     mov        ecx,256
     fninit
 
     fld        .counter
  @@:
     fld        st
     fsincos
     fstp       dword [edi]
     fstp       dword [esi]
;     fadd       [piD180]
     fadd       [piD128]
     add        esi,4
     add        edi,4
     loop       @b
     ffree      st
 
     mov        esp,ebp
     pop        ebp
ret
;------
; esi - offset (pointer) to angles, edi offset to 3x3 matrix
make_rotation_matrix:
   .sinx   equ dword[ebp-4]
   .cosx   equ dword[ebp-8]
   .siny   equ dword[ebp-12]
   .cosy   equ dword[ebp-16]
   .sinz   equ dword[ebp-20]
   .cosz   equ dword[ebp-24]
     push      ebp
     mov       ebp,esp
     sub       esp,24
 
     movzx     ebx,word[esi]
     shl       ebx,2
     mov       eax,dword[sin_tab+ebx]
     mov       .sinx,eax
     mov       edx,dword[cos_tab+ebx]
     mov       .cosx,edx
 
     movzx     ebx,word[esi+2]
     shl       ebx,2
     mov       eax,dword[sin_tab+ebx]
     mov       .siny,eax
     mov       edx,dword[cos_tab+ebx]
     mov       .cosy,edx
 
     movzx     ebx,word[esi+4]
     shl       ebx,2
     mov       eax,dword[sin_tab+ebx]
     mov       .sinz,eax
     mov       edx,dword[cos_tab+ebx]
     mov       .cosz,edx
 
     fninit
     fld       .cosy
     fmul      .cosz
     fstp      dword[edi]
 
     fld       .sinx
     fmul      .siny
     fmul      .cosz
     fld       .cosx
     fmul      .sinz
     fchs
     faddp
     fstp      dword[edi+12]
 
     fld       .cosx
     fmul      .siny
     fmul      .cosz
     fld       .sinx
     fmul      .sinz
     faddp
     fstp      dword[edi+24]
 
     fld       .cosy
     fmul      .sinz
     fstp      dword[edi+4]
 
     fld       .sinx
     fmul      .siny
     fmul      .sinz
     fld       .cosx
     fmul      .cosz
     faddp
     fstp      dword[edi+16]
 
     fld       .cosx
     fmul      .siny
     fmul      .sinz
     fld       .sinx
     fchs
     fmul      .cosz
     faddp
     fstp      dword[edi+28]
 
     fld       .siny
     fchs
     fstp      dword[edi+8]
 
     fld       .cosy
     fmul      .sinx
     fstp      dword[edi+20]
 
     fld       .cosx
     fmul      .cosy
     fstp      dword[edi+32]
 
     mov       esp,ebp
     pop       ebp
ret
;---------------------
;  in:  esi - ptr to points(normals], each point(normal) coeficient as dword
;       edi - ptr to rotated points(normals)
;       ebx - ptr to 3x3 (9 dwords, 36 bytes) rotation matrix
;       ecx - number of points(normals)
rotary:
if Ext
    fninit
 .again:
 
    fld     dword[esi]
    fmul    dword[ebx]
    fld     dword[esi+4]
    fmul    dword[ebx+12]
    faddp
    fld     dword[esi+8]
    fmul    dword[ebx+24]
    faddp
    fstp    dword[edi]
 
 
    fld     dword[esi+4]
    fmul    dword[ebx+16]
    fld     dword[esi]
    fmul    dword[ebx+4]
    faddp
    fld     dword[esi+8]
    fmul    dword[ebx+28]
    faddp
    fstp    dword[edi+4]
 
 
    fld     dword[esi+8]
    fmul    dword[ebx+32]
    fld     dword[esi]
    fmul    dword[ebx+8]
    fld     dword[esi+4]
    fmul    dword[ebx+20]
    faddp
    faddp
    fstp    dword[edi+8]
 
 
    add     esi,12
    add     edi,12
    loop    .again
    mov     [edi],dword -1
else
;   Copyright (C) 1999-2001  Brian Paul
;   Copyright (C)            Maciej Guba
;---------------------
;  in:  esi - ptr to points(normals], each point(normal) coeficient as dword
;       edi - ptr to rotated points(normals)
;       ebx - ptr to 3x3 (9 dwords, 36 bytes) rotation matrix
;       ecx - number of points(normals)
;align 32
    movups   xmm4,[ebx]
 ;   lddqu    xmm4,[ebx]   ; I tried sse3 :D
    movups   xmm5,[ebx+12]
    movups   xmm6,[ebx+24]
;align 32
  .again:
    movss    xmm0,dword[esi]
    shufps   xmm0,xmm0,0
    mulps    xmm0,xmm4
 
    movss    xmm1,dword[esi+4]
    shufps   xmm1,xmm1,0
    mulps    xmm1,xmm5
 
    movss    xmm2,dword[esi+8]
    shufps   xmm2,xmm2,0
    mulps    xmm2,xmm6
 
    addps    xmm0,xmm1
    addps    xmm0,xmm2
 
    movups   [edi],xmm0
 
    add      esi,12
    add      edi,12
    dec      ecx
    jne      .again
    mov      [edi],dword -1
end if
ret
;----------------------------------------------
;  esi - pointer to 3x3 matrix
add_scale_to_matrix:
     fninit
     fld     [rsscale]
     fld     dword[esi]            ;-----
     fmul    st,st1
     fstp    dword[esi]
     fld     dword[esi+12]           ; x scale
     fmul    st,st1
     fstp    dword[esi+12]
     fld     dword[esi+24]
     fmul    st,st1
     fstp    dword[esi+24]         ;------
 
     fld     dword[esi+4]          ;-----
     fmul    st,st1
     fstp    dword[esi+4]
     fld     dword[esi+16]            ; y scale
     fmul    st,st1
     fstp    dword[esi+16]
     fld     dword[esi+28]
     fmul    st,st1
     fstp    dword[esi+28]         ;------
 
 
     fld     dword[esi+8]          ;-----
     fmul    st,st1
     fstp    dword[esi+8]
     fld     dword[esi+20]              ; z scale
     fmul    st,st1
     fstp    dword[esi+20]
     fld     dword[esi+32]
     fmulp    st1,st
     fstp    dword[esi+32]         ;------
 
ret
 
;in   esi - offset to 3d points  (point as 3 dwords float)
;     edi - offset to 2d points  ( as 3 words integer)
;     ecx - number of points
translate_points:  ; just convert into integer; z coord still needed
    fninit
  .again:
  if 0
    fld    dword[esi+8]
 ;   fmul   [rsscale]
    fist   word[edi+4]
 
    fisub  [zobs]
    fchs
 
    fld    dword[esi]
;    fmul   [rsscale]
    fisub  [xobs]
    fimul  [zobs]
    fdiv   st0,st1
 
    fiadd  [xobs]
    fiadd  [vect_x]
    fistp  word[edi]
 
    fld    dword[esi+4]
;    fmul   [rsscale]
    fisub  [yobs]
    fimul  [zobs]
    fdivrp  ;   st0,st1
 
    fiadd  [yobs]
    fiadd  [vect_y]
    fistp  word[edi+2]
   end if
   ; movups   xmm0,[esi]
   ; cvtps2dq xmm0,xmm0
   ; packsdw xmm0,xmm0
   ; movq     [edi]
    fld    dword[esi]
    fiadd  word[vect_x]
    fistp  word[edi]
    fld    dword[esi+4]
    fiadd  [vect_y]
    fistp  word[edi+2]
    fld    dword[esi+8]
    fistp  word[edi+4]
    add    esi,12
    add    edi,6
    dec    ecx
    jnz    .again
 
ret
>

Rev 8719	Rev 9237
1	x3d equ 0	1	x3d equ 0
2	y3d equ 2	2	y3d equ 2
3	z3d equ 4	3	z3d equ 4
4	vec_x equ 0	4	vec_x equ 0
5	vec_y equ 4	5	vec_y equ 4
6	vec_z equ 8	6	vec_z equ 8
-		7
-		8	if 0 ; Ext >= SSE3
-		9	calc_bounding_box:
-		10	; in:
-		11	; xmm0 - normal vector of ray
-		12	; xmm1 - light origin
-		13	; out:
7	; 3d point - triple integer word coordinate	14	; eax - axis aligned bounding boxes bit mask
-		15
-		16	.rmx equ [ebp-36]
-		17	.nray equ [ebp-64]
-		18	.origin equ [ebp-80]
-		19	.dirfrac equ [ebp-96]
-		20	.nrayr equ [ebp-112]
-		21	.originr equ [ebp-128]
-		22	.tmin equ [ebp-132]
-		23	.tmax equ [ebp-136]
-		24
-		25
-		26	push ebp
-		27	mov ebp,esp
-		28	and ebp,-16
-		29	sub esp,160
-		30
-		31	movss xmm5,[rsscale]
-		32	shufps xmm5,xmm1,0
-		33	movd xmm2,[vect_x]
-		34	punpcklwd xmm2,[the_zero]
-		35	cvtdq2ps xmm2,xmm2
-		36	subps xmm1,xmm2
-		37	movaps .origin,xmm1
-		38	mulps xmm0,xmm5
-		39	movaps .nray,xmm0
-		40
-		41	mov esi,matrix
-		42	lea edi,.rmx
-		43	call reverse_mx_3x3
-		44
-		45	; in: esi - ptr to points(normals], each point(normal) coeficient as dword
8	; vector - triple float dword coordinate	46	; edi - ptr to rotated points(normals)
9	;----------------------in: --------------------------------	47	; ebx - ptr to 3x3 (9 dwords, 36 bytes) rotation matrix
10	;------------------------ esi - pointer to 1st 3d point ---	48	; ecx - number of points(normals)
-		49
-		50	; reverse transform
-		51	lea esi,.nray
-		52	lea edi,.nrayr
-		53	lea ebx,.rmx
-		54	mov ecx,1
-		55	call rotary
-		56
-		57	lea esi,.origin
-		58	lea edi,.originr
-		59	lea ebx,.rmx
-		60	mov ecx,1
-		61	call rotary
-		62
-		63	xor ecx,ecx
-		64	mov ebx,aabb1
-		65	xor eax,eax
-		66	rcpps xmm7,.nrayr
-		67	movaps .dirfrac,xmm7
-		68
-		69	.nx_aabb:
-		70	movaps xmm5,[ebx]
-		71	movaps xmm6,[ebx]
-		72	minps xmm5,[the_zero]
-		73	maxps xmm6,[the_zero]
11	;------------------------ edi - pointer to 2nd 3d point ---	74	; xmm5 - lb corner of AABB with minimal coordinates
12	;------------------------ ebx - pointer to result vector --	75	; xmm6 - rt cor. of AABB wit maximum coords
-		76	subps xmm5,.originr
-		77	subps xmm6,.originr
-		78	mulps xmm5,.dirfrac ; xmm5 - tx1, ty1
13	;---------------------- out : none ------------------------	79	mulps xmm6,.dirfrac ; xmm6 - tx2, ty2
-		80	movaps xmm1,xmm6
-		81	movaps xmm2,xmm6
-		82
-		83
-		84	minps xmm1,xmm5
-		85	maxps xmm2,xmm5
-		86
-		87	movaps xmm5,xmm1
-		88	movaps xmm6,xmm2
-		89	shufps xmm5,xmm5,11100001b
-		90	shufps xmm6,xmm6,11100001b
-		91	maxss xmm1,xmm5 ;t min
-		92	minss xmm2,xmm6 ;t max
-		93	comiss xmm2,xmm1
-		94	jb .no_inter
-		95	.yes:
-		96	bts eax,ecx
-		97	.no_inter:
-		98	add ebx,16
-		99	inc ecx
-		100	cmp ecx,8
-		101	jne .nx_aabb
-		102
-		103	; out: eax - bit mask
-		104	add esp,160
-		105	pop ebp
-		106	ret
-		107	end if
-		108
14	reverse_mx_3x3:	109	reverse_mx_3x3:
15	; esi - source matrix	110	; esi - source matrix
16	; edi - desired reversed matrix	111	; edi - desired reversed matrix
17		112
18	push ebp	113	push ebp
19	mov ebp,esp	114	mov ebp,esp
20	sub esp,4	115	sub esp,4
21	.det equ ebp-4	116	.det equ ebp-4
22		117
23	fninit	118	fninit
24	fld dword[esi]	119	fld dword[esi]
25	fmul dword[esi+16]	120	fmul dword[esi+16]
26	fmul dword[esi+32]	121	fmul dword[esi+32]
27	fld dword[esi+12]	122	fld dword[esi+12]
28	fmul dword[esi+28]	123	fmul dword[esi+28]
29	fmul dword[esi+8]	124	fmul dword[esi+8]
30	faddp	125	faddp
31	fld dword[esi+24]	126	fld dword[esi+24]
32	fmul dword[esi+4]	127	fmul dword[esi+4]
33	fmul dword[esi+20]	128	fmul dword[esi+20]
34	faddp	129	faddp
35	fld dword[esi]	130	fld dword[esi]
36	fmul dword[esi+28]	131	fmul dword[esi+28]
37	fmul dword[esi+20]	132	fmul dword[esi+20]
38	fchs	133	fchs
39	faddp	134	faddp
40	fld dword[esi+24]	135	fld dword[esi+24]
41	fmul dword[esi+16]	136	fmul dword[esi+16]
42	fmul dword[esi+8]	137	fmul dword[esi+8]
43	fchs	138	fchs
44	faddp	139	faddp
45	fld dword[esi+12]	140	fld dword[esi+12]
46	fmul dword[esi+4]	141	fmul dword[esi+4]
47	fmul dword[esi+32]	142	fmul dword[esi+32]
48	fchs	143	fchs
49	faddp	144	faddp
50	fstp dword[.det]	145	fstp dword[.det]
51	cmp dword[.det],0	146	cmp dword[.det],0
52	jne @f	147	jne @f
53	int3	148	int3
54	@@:	149	@@:
55	; fld1	150	; fld1
56	; fdiv dword[.det]	151	; fdiv dword[.det]
57	; fstp dword[.det]	152	; fstp dword[.det]
58		153
59	fld dword[esi+16]	154	fld dword[esi+16]
60	fmul dword[esi+32]	155	fmul dword[esi+32]
61	fld dword[esi+20]	156	fld dword[esi+20]
62	fmul dword[esi+28]	157	fmul dword[esi+28]
63	fchs	158	fchs
64	faddp	159	faddp
65	fdiv dword[.det]	160	fdiv dword[.det]
66	fstp dword[edi]	161	fstp dword[edi]
67		162
68	fld dword[esi+8]	163	fld dword[esi+8]
69	fmul dword[esi+28]	164	fmul dword[esi+28]
70	fld dword[esi+4]	165	fld dword[esi+4]
71	fmul dword[esi+32]	166	fmul dword[esi+32]
72	fchs	167	fchs
73	faddp	168	faddp
74	fdiv dword[.det]	169	fdiv dword[.det]
75	fstp dword[edi+4]	170	fstp dword[edi+4]
76		171
77	fld dword[esi+4]	172	fld dword[esi+4]
78	fmul dword[esi+20]	173	fmul dword[esi+20]
79	fld dword[esi+8]	174	fld dword[esi+8]
80	fmul dword[esi+16]	175	fmul dword[esi+16]
81	fchs	176	fchs
82	faddp	177	faddp
83	fdiv dword[.det]	178	fdiv dword[.det]
84	fstp dword[edi+8]	179	fstp dword[edi+8]
85		180
86	fld dword[esi+20]	181	fld dword[esi+20]
87	fmul dword[esi+24]	182	fmul dword[esi+24]
88	fld dword[esi+12]	183	fld dword[esi+12]
89	fmul dword[esi+32]	184	fmul dword[esi+32]
90	fchs	185	fchs
91	faddp	186	faddp
92	fdiv dword[.det]	187	fdiv dword[.det]
93	fstp dword[edi+12]	188	fstp dword[edi+12]
94		189
95	fld dword[esi]	190	fld dword[esi]
96	fmul dword[esi+32]	191	fmul dword[esi+32]
97	fld dword[esi+8]	192	fld dword[esi+8]
98	fmul dword[esi+24]	193	fmul dword[esi+24]
99	fchs	194	fchs
100	faddp	195	faddp
101	fdiv dword[.det]	196	fdiv dword[.det]
102	fstp dword[edi+16]	197	fstp dword[edi+16]
103		198
104	fld dword[esi+8]	199	fld dword[esi+8]
105	fmul dword[esi+12]	200	fmul dword[esi+12]
106	fld dword[esi]	201	fld dword[esi]
107	fmul dword[esi+20]	202	fmul dword[esi+20]
108	fchs	203	fchs
109	faddp	204	faddp
110	fdiv dword[.det]	205	fdiv dword[.det]
111	fstp dword[edi+20]	206	fstp dword[edi+20]
112		207
113	fld dword[esi+12]	208	fld dword[esi+12]
114	fmul dword[esi+28]	209	fmul dword[esi+28]
115	fld dword[esi+16]	210	fld dword[esi+16]
116	fmul dword[esi+24]	211	fmul dword[esi+24]
117	fchs	212	fchs
118	faddp	213	faddp
119	fdiv dword[.det]	214	fdiv dword[.det]
120	fstp dword[edi+24]	215	fstp dword[edi+24]
121		216
122	fld dword[esi+4]	217	fld dword[esi+4]
123	fmul dword[esi+24]	218	fmul dword[esi+24]
124	fld dword[esi]	219	fld dword[esi]
125	fmul dword[esi+28]	220	fmul dword[esi+28]
126	fchs	221	fchs
127	faddp	222	faddp
128	fdiv dword[.det]	223	fdiv dword[.det]
129	fstp dword[edi+28]	224	fstp dword[edi+28]
130		225
131	fld dword[esi]	226	fld dword[esi]
132	fmul dword[esi+16]	227	fmul dword[esi+16]
133	fld dword[esi+4]	228	fld dword[esi+4]
134	fmul dword[esi+12]	229	fmul dword[esi+12]
135	fchs	230	fchs
136	faddp	231	faddp
137	fdiv dword[.det]	232	fdiv dword[.det]
138	fstp dword[edi+32]	233	fstp dword[edi+32]
139		234
140		235
141	mov esp,ebp	236	mov esp,ebp
142	pop ebp	237	pop ebp
143	ret	238	ret
-		239	; 3d point - triple integer word coordinate
-		240	; vector - triple float dword coordinate
-		241	;----------------------in: --------------------------------
-		242	;------------------------ esi - pointer to 1st 3d point ---
-		243	;------------------------ edi - pointer to 2nd 3d point ---
-		244	;------------------------ ebx - pointer to result vector --
-		245	;---------------------- out : none ------------------------
144		246
145	make_vector_r:	247	make_vector_r:
146	if Ext < SSE2	248	if Ext < SSE2
147	fninit	249	fninit
148	fld dword[edi] ;edi+x3d	250	fld dword[edi] ;edi+x3d
149	fsub dword[esi] ;esi+x3d	251	fsub dword[esi] ;esi+x3d
150	fstp dword[ebx+vec_x]	252	fstp dword[ebx+vec_x]
151		253
152	fld dword[edi+4]	254	fld dword[edi+4]
153	fsub dword[esi+4]	255	fsub dword[esi+4]
154	fstp dword[ebx+vec_y]	256	fstp dword[ebx+vec_y]
155		257
156	fld dword[edi+8]	258	fld dword[edi+8]
157	fsub dword[esi+8]	259	fsub dword[esi+8]
158	fstp dword[ebx+vec_z]	260	fstp dword[ebx+vec_z]
159	else	261	else
160	movups xmm0,[esi]	262	movups xmm0,[esi]
161	movups xmm1,[edi]	263	movups xmm1,[edi]
162	subps xmm1,xmm0	264	subps xmm1,xmm0
163	movlps [ebx],xmm1	265	movlps [ebx],xmm1
164	movhlps xmm1,xmm1	266	movhlps xmm1,xmm1
165	movss [ebx+8],xmm1	267	movss [ebx+8],xmm1
166	end if	268	end if
167		269
168	ret	270	ret
169	;---------------------- in: -------------------------------	271	;---------------------- in: -------------------------------
170	;--------------------------- esi - pointer to 1st vector --	272	;--------------------------- esi - pointer to 1st vector --
171	;--------------------------- edi - pointer to 2nd vector --	273	;--------------------------- edi - pointer to 2nd vector --
172	;--------------------------- ebx - pointer to result vector	274	;--------------------------- ebx - pointer to result vector
173	;---------------------- out : none	275	;---------------------- out : none
174	cross_product:	276	cross_product:
175	fninit	277	fninit
176	fld dword [esi+vec_y]	278	fld dword [esi+vec_y]
177	fmul dword [edi+vec_z]	279	fmul dword [edi+vec_z]
178	fld dword [esi+vec_z]	280	fld dword [esi+vec_z]
179	fmul dword [edi+vec_y]	281	fmul dword [edi+vec_y]
180	fsubp ;st1 ,st	282	fsubp ;st1 ,st
181	fstp dword [ebx+vec_x]	283	fstp dword [ebx+vec_x]
182		284
183	fld dword [esi+vec_z]	285	fld dword [esi+vec_z]
184	fmul dword [edi+vec_x]	286	fmul dword [edi+vec_x]
185	fld dword [esi+vec_x]	287	fld dword [esi+vec_x]
186	fmul dword [edi+vec_z]	288	fmul dword [edi+vec_z]
187	fsubp ;st1 ,st	289	fsubp ;st1 ,st
188	fstp dword [ebx+vec_y]	290	fstp dword [ebx+vec_y]
189		291
190	fld dword [esi+vec_x]	292	fld dword [esi+vec_x]
191	fmul dword [edi+vec_y]	293	fmul dword [edi+vec_y]
192	fld dword [esi+vec_y]	294	fld dword [esi+vec_y]
193	fmul dword [edi+vec_x]	295	fmul dword [edi+vec_x]
194	fsubp ;st1 ,st	296	fsubp ;st1 ,st
195	fstp dword [ebx+vec_z]	297	fstp dword [ebx+vec_z]
196	ret	298	ret
-		299	cross_aligned:
-		300	movaps xmm0,[esi]
-		301	movaps xmm1,[esi]
-		302	movaps xmm2,[edi]
-		303	movaps xmm3,[edi]
-		304	shufps xmm0,xmm0,00001001b
-		305	shufps xmm1,xmm1,00010010b
-		306	shufps xmm2,xmm2,00010010b
-		307	shufps xmm3,xmm3,00001001b
-		308	mulps xmm0,xmm2
-		309	mulps xmm1,xmm3
-		310	subps xmm0,xmm1
-		311	movaps [ebx],xmm0
-		312	ret
197	;----------------------- in: ------------------------------	313	;----------------------- in: ------------------------------
198	;---------------------------- edi - pointer to vector -----	314	;---------------------------- edi - pointer to vector -----
199	;----------------------- out : none	315	;----------------------- out : none
200	normalize_vector:	316	normalize_vector:
201	if Ext >= SSE3	317	if Ext >= SSE2
202	movups xmm0,[edi]	318	movups xmm0,[edi]
203	andps xmm0,[zero_hgst_dd]	319	andps xmm0,[zero_hgst_dd]
204	movups xmm1,xmm0	320	movups xmm1,xmm0
205	mulps xmm0,xmm0	321	mulps xmm0,xmm0
-		322	movhlps xmm2,xmm0
-		323	addps xmm0,xmm2
-		324	movaps xmm2,xmm0
-		325	shufps xmm2,xmm2,11100101b
-		326	addps xmm0,xmm2
-		327	shufps xmm0,xmm0,0
206	haddps xmm0,xmm0	328	; haddps xmm0,xmm0
207	haddps xmm0,xmm0	329	; haddps xmm0,xmm0
208	rsqrtps xmm0,xmm0	330	rsqrtps xmm0,xmm0
209	mulps xmm0,xmm1	331	mulps xmm0,xmm1
210	movlps [edi],xmm0	332	movlps [edi],xmm0
211	movhlps xmm0,xmm0	333	movhlps xmm0,xmm0
212	movss [edi+8],xmm0	334	movss [edi+8],xmm0
213	else	335	else
214		336
215	fninit	337	fninit
216	fld dword [edi+vec_x]	338	fld dword [edi+vec_x]
217	fmul st, st	339	fmul st, st
218	fld dword [edi+vec_y]	340	fld dword [edi+vec_y]
219	fmul st, st	341	fmul st, st
220	fld dword [edi+vec_z]	342	fld dword [edi+vec_z]
221	fmul st, st	343	fmul st, st
222	faddp st1, st	344	faddp st1, st
223	faddp st1, st	345	faddp st1, st
224	fsqrt	346	fsqrt
225		347
226	ftst	348	ftst
227	fstsw ax	349	fstsw ax
228	sahf	350	sahf
229	jnz @f	351	jnz @f
230		352
231	fst dword [edi+vec_x]	353	fst dword [edi+vec_x]
232	fst dword [edi+vec_y]	354	fst dword [edi+vec_y]
233	fstp dword [edi+vec_z]	355	fstp dword [edi+vec_z]
234	ret	356	ret
235	@@:	357	@@:
236	fld st	358	fld st
237	fld st	359	fld st
238	fdivr dword [edi+vec_x]	360	fdivr dword [edi+vec_x]
239	fstp dword [edi+vec_x]	361	fstp dword [edi+vec_x]
240	fdivr dword [edi+vec_y]	362	fdivr dword [edi+vec_y]
241	fstp dword [edi+vec_y]	363	fstp dword [edi+vec_y]
242	fdivr dword [edi+vec_z]	364	fdivr dword [edi+vec_z]
243	fstp dword [edi+vec_z]	365	fstp dword [edi+vec_z]
244	end if	366	end if
245	ret	367	ret
246	;------------------in: -------------------------	368	;------------------in: -------------------------
247	;------------------ esi - pointer to 1st vector	369	;------------------ esi - pointer to 1st vector
248	;------------------ edi - pointer to 2nd vector	370	;------------------ edi - pointer to 2nd vector
249	;------------------out: ------------------------	371	;------------------out: ------------------------
250	;------------------ st0 - dot-product	372	;------------------ st0 - dot-product
251	dot_product:	373	dot_product:
252	fninit	374	fninit
253	;if Ext >=SSE3	375	;if Ext >=SSE3
254	; movups xmm0,[esi]	376	; movups xmm0,[esi]
255	; movups xmm1,[edi]	377	; movups xmm1,[edi]
256	; andps xmm0,[zero_hgst_dd]	378	; andps xmm0,[zero_hgst_dd]
257	; mulps xmm0,xmm1	379	; mulps xmm0,xmm1
258	; haddps xmm0,xmm0	380	; haddps xmm0,xmm0
259	; haddps xmm0,xmm0	381	; haddps xmm0,xmm0
260	; movss [esp-4],xmm0	382	; movss [esp-4],xmm0
261	; fld dword[esp-4]	383	; fld dword[esp-4]
262	;else	384	;else
263	fld dword [esi+vec_x]	385	fld dword [esi+vec_x]
264	fmul dword [edi+vec_x]	386	fmul dword [edi+vec_x]
265	fld dword [esi+vec_y]	387	fld dword [esi+vec_y]
266	fmul dword [edi+vec_y]	388	fmul dword [edi+vec_y]
267	fld dword [esi+vec_z]	389	fld dword [esi+vec_z]
268	fmul dword [edi+vec_z]	390	fmul dword [edi+vec_z]
269	faddp	391	faddp
270	faddp	392	faddp
271	;end if	393	;end if
272	ret	394	ret
273		395
274	; DOS version Coded by Mikolaj Felix aka Majuma	396	; DOS version Coded by Mikolaj Felix aka Majuma
275	; mfelix@polbox.com	397	; mfelix@polbox.com
276	; www.majuma.xt.pl	398	; www.majuma.xt.pl
277	; into FASM translation by Macgub	399	; into FASM translation by Macgub
278	init_sincos_tab:	400	init_sincos_tab:
279	.counter equ dword [ebp-4] ; cur angle	401	.counter equ dword [ebp-4] ; cur angle
280		402
281	push ebp	403	push ebp
282	mov ebp,esp	404	mov ebp,esp
283		405
284	xor eax,eax	406	xor eax,eax
285	push eax ; init .counter	407	push eax ; init .counter
286	mov edi,cos_tab	408	mov edi,cos_tab
287	mov esi,sin_tab	409	mov esi,sin_tab
288	mov ecx,256	410	mov ecx,256
289	fninit	411	fninit
290		412
291	fld .counter	413	fld .counter
292	@@:	414	@@:
293	fld st	415	fld st
294	fsincos	416	fsincos
295	fstp dword [edi]	417	fstp dword [edi]
296	fstp dword [esi]	418	fstp dword [esi]
297	; fadd [piD180]	419	; fadd [piD180]
298	fadd [piD128]	420	fadd [piD128]
299	add esi,4	421	add esi,4
300	add edi,4	422	add edi,4
301	loop @b	423	loop @b
302	ffree st	424	ffree st
303		425
304	mov esp,ebp	426	mov esp,ebp
305	pop ebp	427	pop ebp
306	ret	428	ret
307	;------	429	;------
308	; esi - offset (pointer) to angles, edi offset to 3x3 matrix	430	; esi - offset (pointer) to angles, edi offset to 3x3 matrix
309	make_rotation_matrix:	431	make_rotation_matrix:
310	.sinx equ dword[ebp-4]	432	.sinx equ dword[ebp-4]
311	.cosx equ dword[ebp-8]	433	.cosx equ dword[ebp-8]
312	.siny equ dword[ebp-12]	434	.siny equ dword[ebp-12]
313	.cosy equ dword[ebp-16]	435	.cosy equ dword[ebp-16]
314	.sinz equ dword[ebp-20]	436	.sinz equ dword[ebp-20]
315	.cosz equ dword[ebp-24]	437	.cosz equ dword[ebp-24]
316	push ebp	438	push ebp
317	mov ebp,esp	439	mov ebp,esp
318	sub esp,24	440	sub esp,24
319		441
320	movzx ebx,word[esi]	442	movzx ebx,word[esi]
321	shl ebx,2	443	shl ebx,2
322	mov eax,dword[sin_tab+ebx]	444	mov eax,dword[sin_tab+ebx]
323	mov .sinx,eax	445	mov .sinx,eax
324	mov edx,dword[cos_tab+ebx]	446	mov edx,dword[cos_tab+ebx]
325	mov .cosx,edx	447	mov .cosx,edx
326		448
327	movzx ebx,word[esi+2]	449	movzx ebx,word[esi+2]
328	shl ebx,2	450	shl ebx,2
329	mov eax,dword[sin_tab+ebx]	451	mov eax,dword[sin_tab+ebx]
330	mov .siny,eax	452	mov .siny,eax
331	mov edx,dword[cos_tab+ebx]	453	mov edx,dword[cos_tab+ebx]
332	mov .cosy,edx	454	mov .cosy,edx
333		455
334	movzx ebx,word[esi+4]	456	movzx ebx,word[esi+4]
335	shl ebx,2	457	shl ebx,2
336	mov eax,dword[sin_tab+ebx]	458	mov eax,dword[sin_tab+ebx]
337	mov .sinz,eax	459	mov .sinz,eax
338	mov edx,dword[cos_tab+ebx]	460	mov edx,dword[cos_tab+ebx]
339	mov .cosz,edx	461	mov .cosz,edx
340		462
341	fninit	463	fninit
342	fld .cosy	464	fld .cosy
343	fmul .cosz	465	fmul .cosz
344	fstp dword[edi]	466	fstp dword[edi]
345		467
346	fld .sinx	468	fld .sinx
347	fmul .siny	469	fmul .siny
348	fmul .cosz	470	fmul .cosz
349	fld .cosx	471	fld .cosx
350	fmul .sinz	472	fmul .sinz
351	fchs	473	fchs
352	faddp	474	faddp
353	fstp dword[edi+12]	475	fstp dword[edi+12]
354		476
355	fld .cosx	477	fld .cosx
356	fmul .siny	478	fmul .siny
357	fmul .cosz	479	fmul .cosz
358	fld .sinx	480	fld .sinx
359	fmul .sinz	481	fmul .sinz
360	faddp	482	faddp
361	fstp dword[edi+24]	483	fstp dword[edi+24]
362		484
363	fld .cosy	485	fld .cosy
364	fmul .sinz	486	fmul .sinz
365	fstp dword[edi+4]	487	fstp dword[edi+4]
366		488
367	fld .sinx	489	fld .sinx
368	fmul .siny	490	fmul .siny
369	fmul .sinz	491	fmul .sinz
370	fld .cosx	492	fld .cosx
371	fmul .cosz	493	fmul .cosz
372	faddp	494	faddp
373	fstp dword[edi+16]	495	fstp dword[edi+16]
374		496
375	fld .cosx	497	fld .cosx
376	fmul .siny	498	fmul .siny
377	fmul .sinz	499	fmul .sinz
378	fld .sinx	500	fld .sinx

Subversion Repositories Kolibri OS

(root)/programs/demos/view3ds/3dmath.inc – Rev 8719 → 9237